The purpose of the study is to determine the effectiveness of muscle strengthening alone in improving a variety of outcomes in women with fibromyalgia (FM). Specifically, the PI proposes to: (1) initiate a 12- week controlled clinical trial of muscle strengthening vs stretching in a group of low fitness women with FM; and to (2) compare the differences in the two groups for producing changes in: a) muscle strength and endurance, b) body composition, c) aerobic fitness, d) functional ability, e) FM symptoms, f) self-efficacy, and g) psychological status. Sixty women with definite FM of five or fewer years duration who are between the ages of 20 and 50 years will be recruited from a University rheumatology practice which includes over 5000 active FM patients and randomized to one of the two study groups. Each groups will meet for 45 minutes three times a week and will be led by a certified fitness expert with experience with FM patients. Ss will be given a group appropriate exercise videotape which can be substituted for up to one class session per week. Pre- and post-intervention measures will be taken within 14 days preceding the intervention and following the intervention and will be done by the principal investigator who will be blinded to group assignment. The primary outcomes are muscle strength and endurance which will be measured by testing isometric and isokinetic strength of knee extensions using a Cybex isokinetic dynamometer. Additional measures include: a) a VAS for rating perception of ease of exercise; b) body weight, height and body mass index using the Quetelet index; c) waist-to-hip circumference ratio; d) aerobic fitness estimated with V02 max using a modified Balke (3.4 mph) treadmill test and a Gould metabolic cart measuring respiratory gas exchange; e) functional status using the FM Impact Questionnaire; f) FM symptoms assessed by the number and degree of tenderness of tender points and a cumulative myalgia score; g) self-efficacy using the Arthritis Self-Efficacy Score; h) psychological status using as adapted Beck Depression Scale, the Beck Anxiety Inventory; and i) quality of life using the Quality of Life Scale. Proposed analyses include analysis of variance using change scores, analysis of covariance, and/or multiple regression analysis.